Something to keep in mind in the coming couple of weeks: we believe from modeling estimates from case data (github.com/midas-network/…) and from phylodynamics (bedford.io/projects/ncov-…) that epidemic doubling time is about 7 days. 1/4
This means that 100 infections will grow to 200 infections after one week and to 400 infections after two weeks. After local transmission has started, “there’ll be twice as many infections in a week” is a good grounding to think about how things will behave in your area. 2/4
Because of the lack of national testing that had been going on, there is likely a backlog of cases to be detected. As this backlog gets cleared, case counts are going to rise quickly. But please remember that reported cases aren’t newly acquired infections. 3/4
If US cases more than double in the next week it’s not because the virus is moving any faster. It’s just because we’ll finally be catching up to it. 4/4
Postscript: as we’ve seen in China, this 7 day doubling can be mitigated against with effective non-pharmaceutical interventions
Reading these replies, I want to clarify one thing. The 7 day doubling is an estimate and has uncertainty around it. It will also be different in different populations with different contact patterns. Certain situations may have faster growth and others slower growth.
